Age-related macular degeneration (AMD) is a debilitating blinding disease of older age that is increasing in prevalence in the Veteran population. There are currently no highly effective therapies for atrophic (dry) AMD, the most common form of this disorder. It is now appreciated that the vitamin A metabolite, all-trans-retinal, likely plays a key role in the pathogenesis of AMD. All-trans-retinal is released from visual pigments (rod and cone holo-opsins) in photoreceptor cells following their activation by light. For sustained vision, 11-cis-retinal must e continuously provided to the photoreceptor cells so that sufficient levels of light-sensitive holo-opsins are maintained. Rod and cone photoreceptors receive 11-cis-retinal from a metabolic pathway known as the visual cycle, which involves enzymes located in the retinal pigment epithelium. There is strong evidence that an alternative, Mller cell-dependent visual cycle also exists in the retina to provide cone photoreceptors with an additional supply of visual chromophore adequate for operation under daytime lighting conditions. Agents that inhibit production of 11-cis-retinal by targeting the classical visual cycle isomerase, RPE65, protect against light-induced damage to the retina by reducing the amount of all-trans-retinal released after light exposure. These agents have been proposed as potential pharmacotherapies for AMD. One inhibitor of the RPE-dependent visual cycle, called emixustat, is currently being evaluated for its ability to slow AMD progression in a phase 3 clinical trial. RPE65 inhibitors hol great promise as effective agents for the treatment of dry AMD, but they also have undesirable side effects on cone photoreceptors and color vision. The origin of these troublesome ocular side effects remains to be elucidated, but they could potentially arise from off-target interaction with components of the Mller cell-dependent visual cycle. Alternatively, because of the abundance of cones over rods in the human macula where AMD pathology occurs it is also conceivable that partial inhibition of the Mller cell-dependent visual cycle could be therapeutically beneficial for AMD treatment. Recently, a candidate retinoid isomerase of the Mller cell-based pathway, called sphingolipid-?4-desaturase (DES1), was identified in a high-throughput expression screen. This Mller cell-localized enzyme, originally characterized as a dihydroceramide desaturase, was shown to catalyze formation of cis retinol isomers from all-trans-retinol in vitro and partially restore retinal function when overexpressed in Rpe65-/- mice. However, the physiological contribution of this enzyme to cone cell function has not yet been studied. This proposal aims to 1) evaluate the physiological contribution of DES1 to cone-mediated vision using knockout mouse models and 2) capitalize on recently acquired high-resolution structural information on the binding mode of emixustat to the RPE65 active site to design novel visual cycle inhibitors with high selectively towards RPE65 and improved toxicity profiles. Retinal structure and function in Des1 conditional knockout mice will be assessed by high resolution imaging and electroretinography. Structure-guided drug design will be used to generate highly selective RPE65 inhibitors for potential use in the treatment of dry AMD.